CaltechAUTHORS
  A Caltech Library Service

Guaranteeing Safety of Learned Perception Modules via Measurement-Robust Control Barrier Functions

Dean, Sarah and Taylor, Andrew J. and Cosner, Ryan K. and Recht, Benjamin and Ames, Aaron D. (2020) Guaranteeing Safety of Learned Perception Modules via Measurement-Robust Control Barrier Functions. . (Unpublished) https://resolver.caltech.edu/CaltechAUTHORS:20201109-140938872

[img] PDF - Submitted Version
See Usage Policy.

757Kb

Use this Persistent URL to link to this item: https://resolver.caltech.edu/CaltechAUTHORS:20201109-140938872

Abstract

Modern nonlinear control theory seeks to develop feedback controllers that endow systems with properties such as safety and stability. The guarantees ensured by these controllers often rely on accurate estimates of the system state for determining control actions. In practice, measurement model uncertainty can lead to error in state estimates that degrades these guarantees. In this paper, we seek to unify techniques from control theory and machine learning to synthesize controllers that achieve safety in the presence of measurement model uncertainty. We define the notion of a Measurement-Robust Control Barrier Function (MR-CBF) as a tool for determining safe control inputs when facing measurement model uncertainty. Furthermore, MR-CBFs are used to inform sampling methodologies for learning-based perception systems and quantify tolerable error in the resulting learned models. We demonstrate the efficacy of MR-CBFs in achieving safety with measurement model uncertainty on a simulated Segway system.


Item Type:Report or Paper (Discussion Paper)
Related URLs:
URLURL TypeDescription
http://arxiv.org/abs/2010.16001arXivDiscussion Paper
ORCID:
AuthorORCID
Taylor, Andrew J.0000-0002-5990-590X
Recht, Benjamin0000-0002-0293-593X
Ames, Aaron D.0000-0003-0848-3177
Additional Information:We thank the anonymous reviewers for helpful feedback, and Andrew Singletary for his work developing the Segway simulation environment. This research is generously supported in part by ONR awards N00014-20-1-2497 and N00014-18-1-2833, NSF CPS award 1931853, and the DARPA Assured Autonomy program (FA8750-18-C-0101), and a gift from Twitter. SD is supported by an NSF Graduate Research Fellowship under Grant No. DGE 1752814. AT is supported by DARPA award HR00111890035.
Funders:
Funding AgencyGrant Number
Office of Naval Research (ONR)N00014-20-1-2497
Office of Naval Research (ONR)N00014-18-1-2833
NSFCMMI-1931853
Defense Advanced Research Projects Agency (DARPA)FA8750-18-C-0101
TwitterUNSPECIFIED
NSF Graduate Research FellowshipDGE-1752814
Defense Advanced Research Projects Agency (DARPA)HR00111890035
Subject Keywords:safety, measurements, learning, perception
Record Number:CaltechAUTHORS:20201109-140938872
Persistent URL:https://resolver.caltech.edu/CaltechAUTHORS:20201109-140938872
Usage Policy:No commercial reproduction, distribution, display or performance rights in this work are provided.
ID Code:106552
Collection:CaltechAUTHORS
Deposited By: George Porter
Deposited On:09 Nov 2020 23:37
Last Modified:09 Nov 2020 23:37

Repository Staff Only: item control page